OF ARTS AND SCIENCES. 



485 



from this iron was obtained a far more perfect octahedron than from 

 any of those with a coarser structure. The smaller sketch shows one 

 face of such an octahedron, twice the original size. This octahedron 

 showed seven perfectly even regular octahedral faces, the eighth face 

 having been hollowed out where it formed part of the crust of the 

 meteorite. 



Butler, Bates County. 



All the faces were polished and etched so that the plates might be 

 followed through their entire extent. On examination, it appeared 

 that the majority of the plates, including even the finest microscoiDic 

 markings, followed the direction of the octahedral faces. But there 

 will be noticed in the sketch of the octahedral face, that, in addition 

 to the usual directions, there appear certain plates which bisect the 

 facial angle of the octahedron. These plates, when followed over an 

 edge on to an adjacent face, were seen to be parallel to an octahedral 

 edge, showing that they must be dodecahedral instead of octahedral. 



Thus it at once appeared that the WidmanstJittiau figures could not 

 be solely characteristic of octahedral structure. Furthermore, the 

 Butler meteorite seemed to stand between well-marked Widmanstiit- 

 tian figures and the finer Ihies discovered by Neumann and shown 

 by him to be parallel to cube edges. Some of the Butler figures are 

 coarse enough to be classed unquestionably as Widmanstiittian, that 

 is, thev show the three varieties of iron distinguished by Reichenbach, 



